Re: ppc soft-float regression

[Rich Felker <dalias@libc.org>]

On Sun, May 17, 2015 at 12:02:19PM +0200, Felix Janda wrote:
> Waldemar Brodkorb wrote:
> > Hi,
> > 
> > as mentioned the days on IRC.
> > 1.1.9 produces a lot of segfaults on Qemu-PPC with Soft-Float.
> > 1.1.8 was fine.
> > 
> > Execute Qemu as this:
> > qemu-system-ppc -nographic -M bamboo -device e1000,netdev=adk0 \
> > -netdev user,id=adk0 -kernel qemu-ppc-initramfs-kernel
> > -initrd qemu-ppc-musl-initramfs
> > 
> > wbx@chrom:~/musl-ppc $ qemu-system-ppc --version
> > QEMU emulator version 2.3.0, Copyright (c) 2003-2008 Fabrice Bellard
> > 
> > The system stopps at boot. strg-d to continue the boot with
> > /etc/init.d/rcS.
> > 
> > Strace is available in /usr/sbin/strace.
> > 
> > Get the kernels and initramfs from here:
> > http://openadk.org/musl-ppc/
> > 
> > Any ideas?
> 
> Command line parsing in busybox seems to be really broken. For example
> 
> mount -t proc proc /proc
> 
> has no effect and
> 
> dmesg | head -n 1
> 
> segfaults. git bisect says
> 
> # first bad commit: [f3ddd173806fd5c60b3f034528ca24542aecc5b9] dynamic linker bootstrap overhaul

Did you observe a regression here yourself, or just with Waldemar's
binaries which you said were broken? If you think this commit really
caused regressions on ppc I'll look into it but I don't want to waste
time doing that if it's not broken.

Rich